Automated generation of personalized mail

1. A method comprising: receiving, by a computing device, input information related to an offered product or service, two or more layouts of printable content related to the offered product or service, demographics of potential recipients, and actions taken by the potential recipients when presented with online content; executing a machine-learning-based algorithm, wherein the machine-learning-based algorithm maps the input information to: a selection of a particular layout from the two or more layouts that is customized for a particular recipient of the potential recipients, and a representation of the offered product or service; and transmitting one or more output files representing the offered product or service, the particular layout, and the particular recipient, wherein reception of the one or more output files can be used by a printing system to print a mailable physical document addressed to the particular recipient, the mailable physical document conforming to the particular layout, and including the representation of the offered product or service.

2. The method of claim 1 , wherein the actions taken by the potential recipients when presented with the online content comprise actions taken in response to viewing an online advertisement.

3. The method of claim 1 , wherein the actions taken by the potential recipients when presented with the online content comprise viewing a web page related to an offering multiple times before purchasing that offering.

4. The method of claim 1 , wherein the actions taken by the potential recipients when presented with the online content relate to search histories or web traffic of the potential recipients.

5. The method of claim 1 , wherein the input information also includes historical transactions or orders made by the potential recipients, and wherein the particular layout is also based on the historical transactions or orders.

6. The method of claim 1 , wherein the input information also includes proposed transactions for the potential recipients, wherein the particular layout incorporates a particular proposed transaction for the particular recipient that involves the offered product or service.

7. The method of claim 1 , wherein the demographics of the potential recipients comprise ages, genders, or geographic locations of the potential recipients.

8. The method of claim 1 , wherein the one or more output files representing the offered product or service comprise a mail ready file, a recommendation file, and a products file, wherein the mail ready file comprises a recipient identification number, a name, and a mailing address of the particular recipient, wherein the recommendation file comprises a recommendation for the offered product or service, and wherein the products file comprises a product identification number and an image associated with the offered product or service.

9. The method of claim 8 , wherein the one or more output files further comprise a proposed transaction file, wherein the recommendation file further comprises a recommendation for a particular proposed transaction for the particular recipient, and wherein the proposed transaction file comprises a proposed transaction identification number associated with the particular proposed transaction.

10. The method of claim 8 , wherein the one or more output files further comprise a message file, wherein the recommendation file further comprises a recommendation for a particular message for the particular recipient, and wherein the message file comprises a message identification number and text associated with the particular message.

11. The method of claim 1 , further comprising: receiving feedback from the printing system relating to the one or more output files, wherein the feedback comprises a file receipt report and an open jobs file, wherein the file receipt report comprises information relating to a total amount of the output files received by the printing system, a status of the output files received by the printing system, and a total amount of the output files rejected by the printing system, and wherein the open jobs file comprises a report of a total number of print jobs that are still pending.

12. The method of claim 1 , further comprising: choosing the machine-learning-based algorithm, from a plurality of machine-learning-based algorithms, based on whether (i) actions taken by the potential recipients are related to the offered product or service, (ii) predictions regarding effectiveness of advertising the offered product or service to the potential recipients are to be made, or (iii) interaction data exists between the potential recipients and the offered product or service.

13. The method of claim 12 , wherein the machine-learning-based algorithm is chosen to be based on a clustering algorithm when the actions taken by the potential recipients related to the offered product or service are not available.

14. The method of claim 12 , wherein the machine-learning-based algorithm is chosen to be based on generalized additive models, gradient boosting machines, regression, or alternative least squares when predictions regarding the effectiveness of advertising the offered product or service to the potential recipients are to be made.

15. The method of claim 12 , wherein the machine-learning-based algorithm is chosen to be based on explicit alternating least squares when any interaction data that exists between the potential recipients and the offered product or service is below a threshold degree of strength.

16. The method of claim 1 , wherein the particular layout comprises non-customizable fields and customizable fields, wherein the printing system populates the customizable fields with a representation of the offered product or service, a proposed transaction, or a message relating to the offered product or service, and wherein the message is based on the demographics of the particular recipient.

17. A non-transitory computer-readable medium having stored therein instructions executable by a processor to cause a computing device to perform operations comprising: receiving input information related to an offered product or service, two or more layouts of printable content related to the offered product or service, demographics of potential recipients, and actions taken by the potential recipients when presented with online content; executing a machine-learning-based algorithm, wherein the machine-learning-based algorithm maps the input information to: a selection of a particular layout from the two or more layouts that is customized for a particular recipient of the potential recipients, and a representation of the offered product or service; and transmitting one or more output files representing the offered product or service, the particular layout, and the particular recipient, wherein reception of the one or more output files can be used by a printing system to print a mailable physical document addressed to the particular recipient, the mailable physical document conforming to the particular layout, and including a representation of the offered product or service.

18. The non-transitory computer-readable medium of claim 17 , wherein the operations further comprise: receiving feedback from the printing system relating to the one or more output files, wherein the feedback comprises a file receipt report and an open jobs file, wherein the file receipt report comprises information relating to a total amount of the output files received by the printing system, a status of the output files received by the printing system, and a total amount of the output files rejected by the printing system, and wherein the open jobs file comprises a report of a total number of print jobs that are still pending.

19. A computing device comprising: a processor; memory; and program instructions, stored in the memory, that upon execution by the processor cause the computing device to perform operations comprising: receiving input information related to an offered product or service, two or more layouts of printable content related to the offered product or service, demographics of potential recipients, and actions taken by the potential recipients when presented with online content; executing a machine-learning-based algorithm, wherein the machine-learning-based algorithm maps the input information to: a selection of a particular layout from the two or more layouts that is customized for a particular recipient of the potential recipients, and a representation of the offered product or service; and transmitting one or more output files representing the offered product or service, the particular layout, and the particular recipient, wherein reception of the one or more output files can be used by a printing system to print a mailable physical document addressed to the particular recipient, the mailable physical document conforming to the particular layout, and including a representation of the offered product or service.

20. The computing device of claim 19 , wherein the operations further comprise: receiving feedback from the printing system relating to the one or more output files, wherein the feedback comprises a file receipt report and an open jobs file, wherein the file receipt report comprises information relating to a total amount of the output files received by the printing system, a status of the output files received by the printing system, and a total amount of the output files rejected by the printing system, and wherein the open jobs file comprises a report of a total number of print jobs that are still pending.